EVEN last evening the glow was very marked. At 5.6 a belt of orange-colour lay on the horizon near the point of sunset, having a breadth of 4°. From this base shot up the bluish-white cone, while on each side (south-west and north-west and skirting the horizon) the sky had a smoky-brown aspect. The whole was overtopped by an arch of a pale smoky-pink hue, the outer circumference of which reached an altitude of about 30° as measured from the horizon. At 5.8 the bluish-white cone had become more intense, and the eastern sky was of a pale green. At 5.12 the bluish-white cone, with brown sides and orange base, was very distinct. At 5.14 I noted a long ellipse of intensely blue sky, the meridian forming its major diameter. On each side, west and east, were areas of bluish-white, the latter apparently being a reflection of the former, and having a base on the eastern horizon of dull greenish-brown. At 5.21 I examined the sunset sky with the spectroscope. The low sun-band was becoming dark; the merest possible vapour shading appeared, detached, to the left of D, the little “c” lines were clear, and B (dry gas) intense. At 5.31 the low sun-band had become much deeper. By 5.38 a secondary glow had appeared: pale lemon extended for some 7°, overtopped for the next 20° by an intense smoky-pink, the opposite sky being a dull green, while that immediately above the horizon south-west and north was brown, and the landscape was tinted with a warm glow. The little “a” band was now intense, so also was the low sun-band, and all trace of vapour effects to the left of D had vanished. At 5.43 the glow was “settling down” and had a total extension of some 14°, 7° of orange and 7° of green. The whole thing finished off with a belt of pale sea-green about 10° in diameter shortly after 6 o'clock.